Project - MovieLens Data Analysis¶

The GroupLens Research Project is a research group in the Department of Computer Science and Engineering at the University of Minnesota. The data is widely used for collaborative filtering and other filtering solutions. However, we will be using this data to act as a means to demonstrate our skill in using Python to “play” with data.

Datasets Information:¶

Data.csv: It contains information of ratings given by the users to a particular movie. Columns: user id, movie id, rating, timestamp
item.csv: File contains information related to the movies and its genre.
Columns: movie id, movie title, release date, unknown, Action, Adventure, Animation, Children’s, Comedy, Crime, Documentary, Drama, Fantasy, Film-Noir, Horror, Musical, Mystery, Romance, Sci-Fi, Thriller, War, Western
user.csv: It contains information of the users who have rated the movies. Columns: user id, age, gender, occupation, zip code

Objective:¶

To implement the techniques learnt as a part of the course.

Learning Outcomes:¶

Exploratory Data Analysis
Visualization using Python
Pandas – groupby, merging

Domain¶

Internet and Entertainment

Note that the project will need you to apply the concepts of groupby and merging extensively.

1. Import the necessary packages - 2.5 marks¶

#Pandas Package
import pandas as pd

#Numpy Package
import numpy as np  


#Seaborn Package
import seaborn as sns   

#Visualization Packages
#MatplotLib
import matplotlib.pyplot as plt  

# Set color code,font scale..
sns.set(color_codes=True,rc={'figure.figsize':(35.0,35.0)},font_scale=1)
%matplotlib inline

2. Read the 3 datasets into dataframes - 2.5 marks¶

#MovieRatings Data by User
MovieRatings_Data_DF =  pd.read_csv('Data.csv',index_col='movie id')
#MovieRatings Item By Gener
MovieRating_Item_DF  =  pd.read_csv('item.csv',index_col='movie id')
#MovieRating Info By User Age/Gender/Position
MovieRating_User_DF  =  pd.read_csv('user.csv',index_col='user id')

3. Apply info, shape, describe, and find the number of missing values in the data - 5 marks¶

Note that you will need to do it for all the three datasets seperately

#Can Check if it there is any Null Values with below way
# #MovieRatings_Data_DF.isnull().values.any()

#INFO_ABOUT MovieRatings_Data
print(MovieRatings_Data_DF.info())

#SHAPE
print(MovieRatings_Data_DF.shape)

#Describe-FIVE POINTS SUMMARY
print(MovieRatings_Data_DF.describe())


#Data.csv
print(MovieRatings_Data_DF.isnull().any())

print("Data.csv-Number Of Missing Values-->:") 

print(MovieRatings_Data_DF.isnull().sum().sum())

<class 'pandas.core.frame.DataFrame'>
Int64Index: 100000 entries, 242 to 203
Data columns (total 3 columns):
 #   Column     Non-Null Count   Dtype
---  ------     --------------   -----
 0   user id    100000 non-null  int64
 1   rating     100000 non-null  int64
 2   timestamp  100000 non-null  int64
dtypes: int64(3)
memory usage: 3.1 MB
None
(100000, 3)
            user id         rating     timestamp
count  100000.00000  100000.000000  1.000000e+05
mean      462.48475       3.529860  8.835289e+08
std       266.61442       1.125674  5.343856e+06
min         1.00000       1.000000  8.747247e+08
25%       254.00000       3.000000  8.794487e+08
50%       447.00000       4.000000  8.828269e+08
75%       682.00000       4.000000  8.882600e+08
max       943.00000       5.000000  8.932866e+08
user id      False
rating       False
timestamp    False
dtype: bool
Data.csv-Number Of Missing Values-->:
0

#Can Check if it there is any Null Values with below way
# #MovieRatings_Item.isnull().values.any()

#INFO_ABOUT MovieRatings_Data
print(MovieRating_Item_DF.info())

#SHAPE
print(MovieRating_Item_DF.shape)

#Describe-FIVE POINTS SUMMARY
print(MovieRating_Item_DF.describe())


#Data.csv
print(MovieRating_Item_DF.isnull().any())

print("Item.csv-Number of Missing Values-->:") 
print(MovieRating_Item_DF.isnull().sum().sum())

<class 'pandas.core.frame.DataFrame'>
Int64Index: 1681 entries, 1 to 1682
Data columns (total 21 columns):
 #   Column        Non-Null Count  Dtype 
---  ------        --------------  ----- 
 0   movie title   1681 non-null   object
 1   release date  1681 non-null   object
 2   unknown       1681 non-null   int64 
 3   Action        1681 non-null   int64 
 4   Adventure     1681 non-null   int64 
 5   Animation     1681 non-null   int64 
 6   Childrens     1681 non-null   int64 
 7   Comedy        1681 non-null   int64 
 8   Crime         1681 non-null   int64 
 9   Documentary   1681 non-null   int64 
 10  Drama         1681 non-null   int64 
 11  Fantasy       1681 non-null   int64 
 12  Film-Noir     1681 non-null   int64 
 13  Horror        1681 non-null   int64 
 14  Musical       1681 non-null   int64 
 15  Mystery       1681 non-null   int64 
 16  Romance       1681 non-null   int64 
 17  Sci-Fi        1681 non-null   int64 
 18  Thriller      1681 non-null   int64 
 19  War           1681 non-null   int64 
 20  Western       1681 non-null   int64 
dtypes: int64(19), object(2)
memory usage: 288.9+ KB
None
(1681, 21)
           unknown       Action    Adventure    Animation    Childrens  \
count  1681.000000  1681.000000  1681.000000  1681.000000  1681.000000   
mean      0.000595     0.149316     0.080309     0.024985     0.072576   
std       0.024390     0.356506     0.271852     0.156126     0.259516   
min       0.000000     0.000000     0.000000     0.000000     0.000000   
25%       0.000000     0.000000     0.000000     0.000000     0.000000   
50%       0.000000     0.000000     0.000000     0.000000     0.000000   
75%       0.000000     0.000000     0.000000     0.000000     0.000000   
max       1.000000     1.000000     1.000000     1.000000     1.000000   

            Comedy        Crime  Documentary        Drama      Fantasy  \
count  1681.000000  1681.000000  1681.000000  1681.000000  1681.000000   
mean      0.300416     0.064842     0.029744     0.431291     0.013087   
std       0.458576     0.246321     0.169931     0.495404     0.113683   
min       0.000000     0.000000     0.000000     0.000000     0.000000   
25%       0.000000     0.000000     0.000000     0.000000     0.000000   
50%       0.000000     0.000000     0.000000     0.000000     0.000000   
75%       1.000000     0.000000     0.000000     1.000000     0.000000   
max       1.000000     1.000000     1.000000     1.000000     1.000000   

         Film-Noir       Horror      Musical      Mystery      Romance  \
count  1681.000000  1681.000000  1681.000000  1681.000000  1681.000000   
mean      0.014277     0.054729     0.033314     0.036288     0.146936   
std       0.118667     0.227519     0.179507     0.187061     0.354148   
min       0.000000     0.000000     0.000000     0.000000     0.000000   
25%       0.000000     0.000000     0.000000     0.000000     0.000000   
50%       0.000000     0.000000     0.000000     0.000000     0.000000   
75%       0.000000     0.000000     0.000000     0.000000     0.000000   
max       1.000000     1.000000     1.000000     1.000000     1.000000   

            Sci-Fi     Thriller          War      Western  
count  1681.000000  1681.000000  1681.000000  1681.000000  
mean      0.060083     0.149316     0.042237     0.016062  
std       0.237712     0.356506     0.201189     0.125751  
min       0.000000     0.000000     0.000000     0.000000  
25%       0.000000     0.000000     0.000000     0.000000  
50%       0.000000     0.000000     0.000000     0.000000  
75%       0.000000     0.000000     0.000000     0.000000  
max       1.000000     1.000000     1.000000     1.000000  
movie title     False
release date    False
unknown         False
Action          False
Adventure       False
Animation       False
Childrens       False
Comedy          False
Crime           False
Documentary     False
Drama           False
Fantasy         False
Film-Noir       False
Horror          False
Musical         False
Mystery         False
Romance         False
Sci-Fi          False
Thriller        False
War             False
Western         False
dtype: bool
Item.csv-Number of Missing Values-->:
0

#Can Check if it there is any Null Values with below way
# #MovieRating_User_DF.isnull().values.any()

#INFO_ABOUT MovieRatings_Data
print(MovieRating_User_DF.info())

#SHAPE
print(MovieRating_User_DF.shape)

#Describe-FIVE POINTS SUMMARY
print(MovieRating_User_DF.describe())


#Data.csv
print(MovieRating_User_DF.isnull().any())

print("User.csv-Null Values Count-->:") 
print(MovieRating_User_DF.isnull().sum().sum())

<class 'pandas.core.frame.DataFrame'>
Int64Index: 943 entries, 1 to 943
Data columns (total 4 columns):
 #   Column      Non-Null Count  Dtype 
---  ------      --------------  ----- 
 0   age         943 non-null    int64 
 1   gender      943 non-null    object
 2   occupation  943 non-null    object
 3   zip code    943 non-null    object
dtypes: int64(1), object(3)
memory usage: 36.8+ KB
None
(943, 4)
              age
count  943.000000
mean    34.051962
std     12.192740
min      7.000000
25%     25.000000
50%     31.000000
75%     43.000000
max     73.000000
age           False
gender        False
occupation    False
zip code      False
dtype: bool
User.csv-Null Values Count-->:
0

4. Find the number of movies per genre using the item data - 2.5 marks¶

# use sum on the default axis

#numeric_only = True==>Get Count only Columns with Integer Values

#SORT Genre By Total Number of Movies
MovieRating_Item_DF.sum(numeric_only = True).reset_index(name ='Number Of Movies Per Genre').sort_values(by='Number Of Movies Per Genre')

# Applying Styles for Count Column
cm = sns.light_palette("green", as_cmap=True)

MovieRating_Item_DF.sum(numeric_only = True).reset_index(name ='Number Of Movies Per Genre').sort_values(by='Number Of Movies Per Genre').style.background_gradient(cmap=cm)

5. Find the movies that have more than one genre - 5 marks¶

#hint: use sum on the axis = 1

#STEP-1 :CLONE CURRENT ITEM FRAME INTO NEW FRAME
MovieRating_Item_Genre_DF=MovieRating_Item_DF.copy()

#STEP-2 : GET SUM BY USING AXIS=1 AND STORE VALUES IN NEW COLUMN NAME "Total Types of Genre"
MovieRating_Item_Genre_DF.loc[:,'Total Types of Genre'] = MovieRating_Item_Genre_DF.sum(axis=1)

#STEP-3

#WE CAN DO BY USING iloc
MovieRating_Item_DF[MovieRating_Item_DF.iloc[:, 3:].sum(1)>1]['movie title']

# OR GET MOVIES INFO BASED ON Total Type Count >1
cm = sns.light_palette("green", as_cmap=True)
Movies_MorethanOneGenre_DF=MovieRating_Item_Genre_DF[MovieRating_Item_Genre_DF['Total Types of Genre'] > 1]
Movies_MorethanOneGenre_DF.sort_values(by='Total Types of Genre').style.background_gradient(cmap=cm)

6. Drop the movie where the genre is unknown - 2.5 marks¶

#ONE METHOD-FILTERING DATA FRAME BY UNKNOWN!=1
MovieRating_Item_DF = MovieRating_Item_DF[MovieRating_Item_DF['unknown']!=1]

#ALTERNATIVE-BY USING DROP METHOD
MovieRating_Item_DF.drop(MovieRating_Item_DF.loc[MovieRating_Item_DF['unknown']==1].index, inplace=True)
  
#Checking Count For Test

               
MovieRating_Item_DF.sum(numeric_only = True).reset_index(name ='Number Of Movies Per Genre').sort_values(by='Number Of Movies Per Genre').style.apply(lambda x: ['background: yellow' if x.name in [0]   else '' for i in x], axis=1)

7. Univariate plots of columns: 'rating', 'Age', 'release year', 'Gender' and 'Occupation' - 10 marks¶

# HINT: use distplot for age and countplot for gender,ratings,occupation, release year.
# HINT: Please refer to the below snippet to understand how to get to release year from release date. You can use str.split()
# as depicted below.

a = 'My*cat*is*brown'
print(a.split('*')[3])

#similarly, the release year needs to be taken out from release date

#also you can simply slice existing string to get the desired data, if we want to take out the colour of the cat

print(a[10:])
print(a[-5:])

brown
brown
brown

#your answers here
#DISTOLOT FOR AGE
sns.distplot(MovieRating_User_DF['age'],rug=True, norm_hist=False)
plt.show()

#COUNRPLOT FOR RATING
sns.countplot(data=MovieRatings_Data_DF, x="rating")
plt.show()

#COUNRPLOT FOR GENDER
sns.set(rc={'figure.figsize':(35.0,35.0)},font_scale=3)
sns.countplot( data=MovieRating_User_DF, x="gender")
plt.show()
sns.countplot( hue="age",data=MovieRating_User_DF, x="gender")
plt.show()

#occupationChart=sns.countplot(MovieRating_User_DF['occupation'])
occupationChart = sns.countplot(data=MovieRating_User_DF,x='occupation')
occupationChart.set_xticklabels(occupationChart.get_xticklabels(), rotation=45, horizontalalignment='right')
plt.show()
occupationChart = sns.countplot(data=MovieRating_User_DF,x='occupation',hue='age')
occupationChart.set_xticklabels(occupationChart.get_xticklabels(), rotation=45, horizontalalignment='right')
plt.show()

#Release Year
MovieRating_Year_DF=MovieRating_Item_DF.copy()
MovieRating_Year_DF.loc[:,'release_year'] = MovieRating_Year_DF['release date'].str.split('-',expand=True)[2]

yearChart = sns.countplot(
    data=MovieRating_Year_DF,
    x='release_year',
    palette='Set1'
)
yearChart.set_yticklabels(yearChart.get_yticklabels(), rotation=65)
plt.xticks(rotation=70)
plt.tight_layout()

8. Visualize how popularity of genres has changed over the years - 10 marks¶

Note that you need to use the number of releases in a year as a parameter of popularity of a genre

Hint

1: you need to reach to a data frame where the release year is the index and the genre is the column names (one cell shows the number of release in a year in one genre) or vice versa. Once that is achieved, you can either use multiple bivariate plots or can use the heatmap to visualise all the changes over the years in one go.

Hint 2: Use groupby on the relevant column and use sum() on the same to find out the nuumber of releases in a year/genre.

#Your answer here
#SETTING INDEX AS RELEASE_YEAR
if 'release_year' in MovieRating_Year_DF.columns:
  MovieRating_Year_DF.set_index('release_year',inplace=True, drop=True)

#SUM AS PER RELEASE YEAR
YearMovies= MovieRating_Year_DF.groupby(['release_year']).sum()

#ALSO ADDING NUMBER OF MOVIER PER YEAR TREND TO analysis
YearMovies['number_of_movies_year']= YearMovies.iloc[:, 1:].sum(axis=1)

YearMovies_all_transposed = YearMovies.transpose()
sns.heatmap(YearMovies_all_transposed,vmin=0, vmax=250,cbar_kws={"orientation": "horizontal"},cmap="YlGnBu")
plt.show()

#DIFFERENT VIEW
sns.heatmap(YearMovies,vmin=0, vmax=250,cbar_kws={"orientation": "horizontal"},cmap="YlGnBu")
plt.show()
YearMovies['index'] = YearMovies.index

# CAN USE SCATTER PLOTS FOR SEE TREND BY YEAR-NOT BY GENRE
sns.scatterplot(x="number_of_movies_year", y="index", data=YearMovies,
                   legend='full' )
plt.show()

9. Find the top 25 movies according to average ratings such that each movie has number of ratings more than 100 - 10 marks¶

Hint :

First find the movies that have more than 100 ratings(use merge, groupby and count). Extract the movie id in a list.
Find the average rating of all the movies and sort them in the descending order. You will have to use the .merge() function to reach to a data set through which you can get the ids and the average rating.
Use isin(list obtained from 1) to filter out the movies which have more than 100 ratings.

Note: This question will need you to research about groupby and apply your findings. You can find more on groupby on https://realpython.com/pandas-groupby/.

#your answer here
#Merge Two Data Frames
movieRatingsFrame_DF=MovieRating_Item_DF.merge(MovieRatings_Data_DF, on='movie id')

#Count Of Ratings
#Aggregate by Rating
movie_stats = movieRatingsFrame_DF.groupby('movie id').agg({'rating': [np.size, np.mean]})

#Get Ratings >100
movieList=movie_stats[movie_stats['rating']['size'] > 100]

#Find Average Rating
ratings_mean_count = pd.DataFrame(movieRatingsFrame_DF.groupby('movie id')['rating'].mean())

#SORT-Values
ratings_mean_count=ratings_mean_count.sort_values('rating',ascending=False)

#TOP25 Movies
top25Movies=ratings_mean_count[ratings_mean_count.index.isin(movieList.index)].sort_values(['rating'], ascending=False)[:25]

top25Movies.style.background_gradient(cmap=cm)

10. See gender distribution across different genres check for the validity of the below statements - 10 marks¶

Men watch more drama than women
Women watch more Sci-Fi than men
Men watch more Romance than women

There is no need to conduct statistical tests around this. Just compare the percentages and comment on the validity of the above statements.
you might want ot use the .sum(), .div() function here.
Use number of ratings to validate the numbers. For example, if out of 4000 ratings received by women, 3000 are for drama, we will assume that 75% of the women watch drama.

#Analysis for Genre-Drama
#Get Drama Genre Movies
rating_user_movie_df_All = MovieRatings_Data_DF.merge(MovieRating_Item_DF,on='movie id')
movie_rating_byGender_All = rating_user_movie_df_All.merge(MovieRating_User_DF, on='user id')

#With Size
movie_rating_Gender_Sum_All= movie_rating_byGender_All.groupby(['gender']).agg({'rating': np.size})

#With Sum
movie_rating_Gender_Sum_All_SUM= movie_rating_byGender_All.groupby(['gender']).agg({'rating': np.sum})

DramMovieFrame=MovieRating_Item_DF[MovieRating_Item_DF['Drama'] > 0]['Drama']
DramMovieFrame
rating_user_movie_df = MovieRatings_Data_DF.merge(DramMovieFrame,on='movie id')
movie_rating_byGender = rating_user_movie_df.merge(MovieRating_User_DF, on='user id')
#Size
movie_rating_Gender_Sum = movie_rating_byGender.groupby(['gender']).agg({'rating': np.size})
#With Sum
movie_rating_Gender_Sum_BySum= movie_rating_byGender.groupby(['gender']).agg({'rating': np.sum})

movie_watch_facts = movie_rating_Gender_Sum_BySum.div(movie_rating_Gender_Sum_All_SUM['rating']/100,axis=0)
movie_watch_facts.rename(columns={"rating": "Watch Pecentage(%) for Drama Genre"}).style.background_gradient(cmap=cm)

movie_rating_Gender_Sum = movie_rating_byGender.groupby(['gender']).agg({'rating': np.size})
movie_watch_facts_Size = movie_rating_Gender_Sum.div(movie_rating_Gender_Sum_All['rating']/100,axis=0)


print(movie_watch_facts)


movie_watch_facts_Size.rename(columns={"rating": "Watch Pecentage(%) for Drama Genre By Size"}).style.background_gradient(cmap=cm)

           rating
gender           
F       44.352763
M       40.751736

#Analysis for Genre-Sci-Fi
#Get Sci-Fi Genre Movies
SciFiMovieFrame=MovieRating_Item_DF[MovieRating_Item_DF['Sci-Fi'] > 0]['Sci-Fi']
SciFiMovieFrame
rating_user_movie_df = MovieRatings_Data_DF.merge(SciFiMovieFrame, on='movie id')
movie_rating_byGender = rating_user_movie_df.merge(MovieRating_User_DF, on='user id')
#Ratings Count
movie_rating_Gender_Sum = movie_rating_byGender.groupby(['gender']).agg({'rating': np.sum})

movie_rating_Gender_Size = movie_rating_byGender.groupby(['gender']).agg({'rating': np.size})


movie_watch_facts =  movie_rating_Gender_Sum.div(movie_rating_Gender_Sum_All_SUM['rating']/100,axis=0)
movie_watch_facts.rename(columns={"rating": "Watch Pecentage(%) for Sci-Fi Genre As per Ratings"}).style.background_gradient(cmap=cm)


movie_watch_facts_size=  movie_rating_Gender_Sum.div(movie_rating_Gender_Sum_All['rating']/100,axis=0)
movie_watch_facts.rename(columns={"rating": "Watch Pecentage(%) for Sci-Fi Genre As per Ratings"}).style.background_gradient(cmap=cm)

print(movie_watch_facts)

movie_watch_facts_size.rename(columns={"rating": "Watch Pecentage(%) for Sci_Fi Genre By Size"}).style.background_gradient(cmap=cm)

           rating
gender           
F       10.117279
M       13.787682

#Analysis for Genre-Romance
#Get Romance Genre Movies
RomanceMovieFrame=MovieRating_Item_DF[MovieRating_Item_DF['Romance'] > 0]['Romance']
RomanceMovieFrame
rating_user_movie_df = MovieRatings_Data_DF.merge(RomanceMovieFrame, on='movie id')
movie_rating_byGender = rating_user_movie_df.merge(MovieRating_User_DF, on='user id')
movie_rating_Gender_Sum = movie_rating_byGender.groupby(['gender']).agg({'rating': np.size})

movie_watch_facts =  movie_rating_Gender_Sum.div(movie_rating_Gender_Sum_All_SUM['rating']/100,axis=0)
movie_watch_facts.rename(columns={"rating": "Watch Pecentage(%) for Sci-Fi Genre As per Ratings"}).style.background_gradient(cmap=cm)


movie_watch_facts_size=  movie_rating_Gender_Sum.div(movie_rating_Gender_Sum_All['rating']/100,axis=0)
movie_watch_facts.rename(columns={"rating": "Watch Pecentage(%) for Sci-Fi Genre As per Ratings"}).style.background_gradient(cmap=cm)

print(movie_watch_facts)

movie_watch_facts_size.rename(columns={"rating": "Watch Pecentage(%) for Romance Genre By Size"}).style.background_gradient(cmap=cm)

          rating
gender          
F       6.444870
M       5.190796

Conclusion:¶

Men watch more drama than women:¶

FALSE

Women watch more Sci-Fi than men¶

FALSE

Men watch more Romance than women¶

FALSE

	rating
movie id
408	4.491071
318	4.466443
169	4.466102
483	4.456790
64	4.445230
603	4.387560
12	4.385768
50	4.358491
178	4.344000
134	4.292929
427	4.292237
357	4.291667
98	4.289744
480	4.284916
127	4.283293
285	4.265432
272	4.262626
657	4.259542
474	4.252577
174	4.252381
479	4.251397
313	4.245714
511	4.231214
484	4.210145
172	4.204360

	index	Number Of Movies Per Genre
0	unknown	1
9	Fantasy	22
10	Film-Noir	24
18	Western	27
3	Animation	42
7	Documentary	50
12	Musical	56
13	Mystery	61
17	War	71
11	Horror	92
15	Sci-Fi	101
6	Crime	109
4	Childrens	122
2	Adventure	135
14	Romance	247
1	Action	251
16	Thriller	251
5	Comedy	505
8	Drama	725

	movie title	release date	unknown	Action	Adventure	Animation	Childrens	Comedy	Crime	Documentary	Drama	Fantasy	Film-Noir	Horror	Musical	Mystery	Romance	Sci-Fi	Thriller	War	Western	Total Types of Genre
movie id
1680	Sliding Doors	01-Jan-1998	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	2
1161	Palookaville	25-Oct-1996	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2
625	Sword in the Stone, The	01-Jan-1963	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2
626	So Dear to My Heart	01-Jan-1949	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2
628	Sleepers	18-Oct-1996	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	2
629	Victor/Victoria	01-Jan-1982	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2
630	Great Race, The	01-Jan-1965	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2
623	Angels in the Outfield	01-Jan-1994	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	2
1160	Love! Valour! Compassion!	16-May-1997	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	2
637	Howling, The	01-Jan-1981	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0	0	0	2
641	Paths of Glory	01-Jan-1957	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0	2
1156	Cyclo	02-Aug-1996	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	2
643	The Innocent	01-Jan-1994	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	2
647	Ran	01-Jan-1985	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0	2
648	Quiet Man, The	01-Jan-1952	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0	0	0	0	2
1159	Stalker	01-Jan-1979	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	2
622	Swiss Family Robinson	01-Jan-1960	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2
619	Extreme Measures	27-Sep-1996	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	2
616	Night of the Living Dead	01-Jan-1968	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	1	0	0	0	2
587	Hour of the Pig, The	01-Jan-1993	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	2
1176	Welcome To Sarajevo	01-Jan-1997	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0	2
1174	Caught	27-Sep-1996	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	2
591	Primal Fear	30-Mar-1996	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	2
592	True Crime	01-Jan-1995	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	0	2
594	Heavy	05-Jun-1996	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	2
1173	Bliss	06-Jun-1997	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	2
597	Eraser	21-Jun-1996	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	2
598	Big Squeeze, The	06-Sep-1996	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	2
601	For Whom the Bell Tolls	01-Jan-1943	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	2
602	American in Paris, An	01-Jan-1951	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	0	0	2
603	Rear Window	01-Jan-1954	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	0	2
607	Rebecca	01-Jan-1940	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	1	0	0	2
1170	Spanking the Monkey	01-Jan-1994	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	2
1165	Big Bully	01-Jan-1996	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	2
1155	Rendezvous in Paris (Rendez-vous de Paris, Les)	28-Jun-1996	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0	0	0	0	2
584	Secret Garden, The	01-Jan-1993	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2
1153	Backbeat	01-Jan-1993	0	0	0	0	0	0	0	0	1	0	0	0	1	0	0	0	0	0	0	2
1152	In Love and War	24-Jan-1997	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1	0	2
695	Kicking and Screaming	01-Jan-1995	0	0	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0	2